Yes it is for a shunt burning it away if not closely demanded by the load. Lets find out what's going on with this build first.
5.2V you now got is the base Vref of two LEDs Vf + two transistors Vbe. Your particular ones. The trimmer is not contributing right now.
When other things like proper CCS Amps are verified as set for, and high dissipation spots are under proper sinking control, you may short/bypass one LED underneath so to kill its Vf contribution and get more downwards leeway with adjusting Vout. To can take it from 5.2V to 5V.
Another way is to make R5 R6 higher value so less current runs through the LEDs to produce 0.1V less Vf each. Or replacing with lower within tolerance Vf LED samples. But those ways are tweaky enough until hitting the right mark. Easiest test is to bypass one LED.
When you set 2A CCS the dissipated Rf power is sqIxR so (2*2)x0.33=1.32W and you want at least 3xdiss for reliability i.e. at least a 5W class resistor. So you did well to upgrade it.
Basic questions are:
-Did you confirm the set current by RfVdrop/RfΩ?
-Are 2A peaks what your application really needs or its some external SMPS rating that is never used to its full and such a high setting is wasted heat?
The second question will be answered by monitoring the application's load current with Ameter in series. Note its current peak, possibly occurring at start up and/or high processing phases, set just 100-150mA CC above that figure. To practice wasted heat economy.
Basic questions are:
-Did you confirm the set current by RfVdrop/RfΩ?
-Are 2A peaks what your application really needs or its some external SMPS rating that is never used to its full and such a high setting is wasted heat?
The second question will be answered by monitoring the application's load current with Ameter in series. Note its current peak, possibly occurring at start up and/or high processing phases, set just 100-150mA CC above that figure. To practice wasted heat economy.
Ok V is holding at 5.1 V now
Shorting 1LED actually increased V to 5.65 so I reversed that
Across Rf V=.58, So .58/.33= 1.76 A
When I box it up I will use a clamp ammeter to measure the draw when connected to the Qutest and adjust R1 to reduce current
I put small heat sinks on the diodes but they’re still hot 50-60C - this tolerable?
Main heat sink is 47C
Shorting 1LED actually increased V to 5.65 so I reversed that
Across Rf V=.58, So .58/.33= 1.76 A
When I box it up I will use a clamp ammeter to measure the draw when connected to the Qutest and adjust R1 to reduce current
I put small heat sinks on the diodes but they’re still hot 50-60C - this tolerable?
Main heat sink is 47C
Hmm shorting one LED and Vout increasing is weird unless the JFET for Q2 Q3 got upset due to low voltage. The other ways mentioned to shave 0.1-0.2V if imperative are still valid.
Such TO-220 diodes are tough but better be sinked more agressively not to radiate that much in a box. Unless you will see less CCS is needed and you reset lower, thus the dissipation will generally go down.
Such TO-220 diodes are tough but better be sinked more agressively not to radiate that much in a box. Unless you will see less CCS is needed and you reset lower, thus the dissipation will generally go down.
I have been making several psu.
I tried l-adapter,ubib,superreg for ACP+.
Ubib and super refulator have noise like booon (probably ham) on back ground.
L-adapter and smps didn't have noise then I
changed cheap rcore (30w 24v) then talema(120w24v).
Talema didn't have noise.
I think cheap rcore is not good enough quality.
I am using same rcore transformer on my dac but there was no problem.
I tried l-adapter,ubib,superreg for ACP+.
Ubib and super refulator have noise like booon (probably ham) on back ground.
L-adapter and smps didn't have noise then I
changed cheap rcore (30w 24v) then talema(120w24v).
Talema didn't have noise.
I think cheap rcore is not good enough quality.
I am using same rcore transformer on my dac but there was no problem.
Attachments
Indeed. Best R-Core directly available in EU was from a French store that closed.
Was SHILCHAR Indian OEM trained and equipped in R-Core by Japanese experts.
Attachments
Thank you for information.
Super refulator from Jan Didden's.
I could trust Japanese phenix rcore but it's expensive.
Antek is not expensive but shipping costs.
I don't understand what cause noise,may be my implementation is not good so
I am checking one by one.
I will use talema for ACP+ this time.
Actually I am making a lot of PSU so I bought cheap transformers.
I noticed nuvotem talema is good.
Unfortunately sound dose not come out yet in my streaming dac project.
Super refulator from Jan Didden's.
I could trust Japanese phenix rcore but it's expensive.
Antek is not expensive but shipping costs.
I don't understand what cause noise,may be my implementation is not good so
I am checking one by one.
I will use talema for ACP+ this time.
Actually I am making a lot of PSU so I bought cheap transformers.
I noticed nuvotem talema is good.
Unfortunately sound dose not come out yet in my streaming dac project.
Attachments
From 30W to 120W there is a no small power difference that could be the reason of your preference.
The R-core have electroestatic wire? If yes try yo not connect to chassis.
Also use short wires, well routed and fixed to chassis.
Yes,120w and 30w is big difference, actually I prefer Rcore sound if there is no noise like "wooooo".I have ordered 50w toroidal transformer,so see it happpens.
May be I might buy better quality Rcore transformer.
No connect electroestatic wire on chassis?
I tried but I did not hear any difference.
Shorter wire yes. When I decided which one to use.
Thank you
May be I might buy better quality Rcore transformer.
No connect electroestatic wire on chassis?
I tried but I did not hear any difference.
Shorter wire yes. When I decided which one to use.
Thank you
Kitamura Kiden we used to see in some Sony, Marantz, etc. are they still in business?
Thank you Salas.
I will check different orientation and Quasimodo snubber.
Or lt4320 rectifier later.
Kitamura Kiden is still running business,they just don't accept personal order.
Phenix accept personal custom order with good quality
Now I have 3 L-adapters on my system.
ONU(12v)-wifi router(19v)-Roon server(19v)
I casually listen from phone using roon arc.
I am satisfied with sound quality.
I will check different orientation and Quasimodo snubber.
Or lt4320 rectifier later.
Kitamura Kiden is still running business,they just don't accept personal order.
Phenix accept personal custom order with good quality
Now I have 3 L-adapters on my system.
ONU(12v)-wifi router(19v)-Roon server(19v)
I casually listen from phone using roon arc.
I am satisfied with sound quality.